Hi,
I tried to run the “from_tensorflow.py” from the tutorials. It ran correctly with “ctx = tvm.gpu(3)” and “tf.device(’/device:GPU:3’)” on all the GPU’s instead of specific one. When I tried to run at specific GPU I gave the command os.environ[“CUDA_VISIBLE_DEVICES”]=“3” it showed an error like
What should I do to make this program run on a specific GPU?